Our overall aim is to develop a precision-based nanotheranostic platform where the imaging signal may serve as an early predictive imaging biomarker for intracellular nanoparticle accumulation and therapeutic response. New anti-cancer agents continue to be developed, but many fail due to the tumor developing (multi-) drug resistance. Cellular membrane proteins acting as a drug efflux pump have been identified, and while some promising agents enter tumor cells, they cannot always be retained long enough to be effective. We aim to exploit the enzyme legumain (an asparaginyl endopeptidase) that is overexpressed in prostate cancer cells for specific cleavage of an olsalazine (Olsa)-conjugated peptide substrate, following which the substrate self- assembles into intracellular nanoparticles. This enzyme-driven self-assembly serves several purposes: 1) intracellular entrapment with minimal drug efflux; 2) prolonged tumor drug exposure; and 3) minimal toxicity to normal organs due to rapid blood clearance of non-assembled single molecules. We have preliminary data demonstrating this concept to be feasible in vivo. Since it does not only serve as an anti-cancer drug through inhibition of DNA methylation, but also as a non-metallic, label-free contrast agent for chemical exchange saturation transfer magnetic resonance imaging (CEST MRI), olsalazine is a unique theranostic agent. The drug can be visualized without modification, allowing direct imaging without pharmacological alterations that may affect self-assembly and/or biodistribution. Following in vitro selection of an optimal Olsa-CBT-800CW-Rn- AAN substrate with maximum tumor cell penetration and retention in legumain-overexpressing DU145 cells (Aim 1), we will test this compound for its in vivo nanotheranostic properties in an orthotopic mouse prostate tumor model (Aim 2) and a transgenic mouse model (TRAMP mouse) where normal prostate cells undergo a malignant transformation over time (Aim 3). If successful, this approach may be extended to other enzyme- targeted CEST MRI-detectable theranostic platforms for imaging tumor aggressiveness, drug accumulation, and predicting therapeutic response.